Aims: Tumors initiated in laboratory rodents by different chemical carcinogens can have distinctive patterns of oncogene and tumor suppressor gene mutation, suggesting that these and other genes may be critical targets of environmental carcinogens. A longstanding area of my research has been to test the hypothesis that environmental exposures produce specific patterns of gene mutation in human tumors. Such patterns can be used both to identify critical target genes and to suggest mutational mechanisms by which an environmental agent causes cancer. If specific carcinogens produce characteristic patterns of gene mutation in tumors, the detection of those patterns would be a powerful tool in studies of environmental risk and for use in prevention and early diagnosis. In a related effort, we have designed a new study to test the hypothesis that exposure correlates with the pattern of mutation in premalignant and normal lung tissues and that such mutations may have prognostic significance for lung cancer development. We are using the Lung Imaging Fluorescent Endoscope (LIFE), a newly developed bronchoscopy technique that is a very sensitive method for detecting premalignant lesions and carcinoma in situ (CIS). CIS and premalignant lesions in the lung are not resected by biopsy and there is no recognized standard-of-care treatment for such lesions. Current practice is to follow them without intervention. This will be the first systematic study to follow prospectively the natural development of lung cancer precursor lesions and CIS over time with sequential biopsies. Detection of molecular defects in such tissues may allow us to identify mutational patterns related to exposure and to provide better estimates of lung cancer risk, progression, and prognosis. Procedures and techniques: Epidemiological studies and PCR, SSCP, sequencing, cloning, and Southern blotting; Lung Imaging Fluorescent Endoscope, bronchoscopy Accomplishments: We examined exposure-specific mutations of ras and p53 in epidemiologic studies of 1) aflatoxin and hepatatis B virus in hepatocellular carcinoma, 2) smoking, occupation, and metabolism gene polymorphisms in bladder cancer, 3) organochlorines and diabetes in pancreatic cancer. We have extended our work on multiple mutations in bladder cancer and find frequent alternative splicing of DNA polymerase b. We have developed sensitive techniques for detecting mutations in tumor tissues and established methods for working with the small and degraded samples often available for epidemiology studies. LIFE Study. We have established a prospective study of people at high risk of developing lung cancer that is designed to test whether molecular changes in normal and preneoplastic bronchial epithelium are correlated with exposure or neoplastic progression. We have established techniques for microdissection and LOH analysis from small biopsy samples and for bronchial epithelial cell